Age Related Maculopathy (ARM) is the leading cause of vision loss in the elderly population in the United States and the Western world and is a major public health issue. Epidemiologic studies have indicated that heredity is a significant risk factor and family studies have further substantiated that ARM could be inherited as a dominant disease with late age of onset and variable expressivity. ARM is not well suited for traditional genetic investigations due to difficulties of clinical ascertainment and the small pedigrees because of its late onset. The evaluation of peripheral fundus for drusen; multiple extramacular drusen (MED) and reticular degeneration of the pigment epithelium (RDPE) has provided improved methods of ascertainment. The Affected Pedigree Member method of genetic analysis of complex inherited traits has successfully determined the linkage of familial Late Onset Alzheimer' s disease and demonstrated the feasibility of studying the genetics of ARM. The development of polymerase chain reaction technique and the closely spaced, polymorphic markers spanning the whole genome will now allow a study of the genetic loci without prior knowledge of candidate gene defects. We propose three specific aims to study the genetics of ARM. Our first aim is to collect affected families in two major population centers. The families will be identified then ascertained using the traditional criteria of vision loss, macular drusen and sub retinal neovascularization, and the new criteria of MED and RDPE. We will evaluate the concordance of the clinical features among family members to define a cohort for genetic study. Our second aim is to find the optimal sample size for recruitment and number of markers needed for genotyping by computer-simulating ARM inheritance. Our third aim is to perform linkage analysis using the Affected Pedigree Member method to identify genetic loci that contribute to ARM susceptibility based on' information obtained with the genotyping of our families. We have already begun by collecting forty five families that suggest autosomal dominant inheritance in thirty- nine pedigrees. The preliminary results of our simulation studies show that it is more powerful to genotype 75 families with markers that are 20 Cm apart than to genotype 150 families at 40 cM distance. To achieve better resolution, the simulation studies show that it is better to genotype another 75 families at 20 cM than to further genotype the original 75 families. This approach illustrates that we can use simulations to select regions for genotyping and analysis with high probability (99%) of success of identifying linkage at reduced cost and effort.